the antennaj of Hepialidfe — Lepidoj^tera Jugatx. 507 



next one ; next two ; four segments have one each ; next 

 two, and so on. Not only is the irregularity in number 

 and position interesting, but the transition from slender 

 bristles to stout ones is also. 



Nowhere in Porina do we find the ventral "sensory" 

 area develop into a definite appendage, but the centring 

 of " sense " hairs on a ventral area which is a feature of 

 Porina antennal structure, suggests that such ventral 

 development might occur, and that we find such in 

 Trictcna is most interesting. 



The female antenna of Tridena {labyrinthica) is remark- 

 able in the appendage — distinctly ventral — being very 

 much larger than the shaft of the segment ; this antenna 

 presents a very near approach to unipectination in the 

 remarkable slender appendage. The swollen area above is, 

 however, the rudiment of lateral pectinations, which we 

 find highly developed in the male, which has probably the 

 most specialized Hepialid antennae. Decidedly the male of 

 Tridena, with its tripectinate antenna, is more h'ghly 

 specialized than the female. We may then regard a 

 laterally-pectinated antenna as specialized, and look back 

 consistently, touching Porina by the way, to the latero- 

 ventral forms — Pielus, bacotii, Gibyra, thence to Hepialus 

 and Oiicoptcra. We suggest divergence in several direc- 

 tions from some such form as Oiicoptcra (but not Oneoptcra 

 itself). It is unlikely there has been convergence from a 

 number of points represented by several forms of pectinated 

 antennae, and some others. 



We can indeed corroborate to some extent the conclu- 

 sions arrived at from a study of the antennae, by examination 

 of the wing neuration. Oiicoptcra, Hepialus, Charagia, 

 Gorgopis, bacotii, Trictcna, are of one pattern ; from a 

 previous paper I make the following quotation^ — " It is 

 as nearly certain as anything of the sort can be that the 

 Hcpiialus- Trictcna pattern of neuration is a generalized 

 pattern, from which the wing patterns of Ilcctomanes, 

 Palpiphorus, and of H'epialiscus, and Porina, have been 

 derived." 



